Flow-through water heaters, which are also referred to as tank-less or instantaneous water heaters, employ one or more chambers which are generally sized only slightly greater than the size of the heating elements, and are capable of instantaneously heating the water to a desired temperature as it flows through the heater. Flow-through water heaters have long been recognized as superior alternatives to conventional water heaters with large storage or holding tanks. The problems with prior art flow-through water heaters include the inability to provide a sufficiently responsive control for the powerful high wattage heating supply necessary to maintain desired temperatures for the wide range of conditions normally associated with domestic and commercial water heating applications. Particularly problematical has been the inability to prevent high temperature overshoot at no flow shutdown resulting from latent heat, as well as the rapid build up of mineral deposits within the water heater in hard water areas.
In order to be commercially successful, flow-through instantaneous or tank-less water heaters must have a plurality of high wattage heating elements and a control scheme which provides the ability to respond very quickly to changes in flow and pressure in order to maintain constant temperature control. In order that one better understands the difficulties associated with both tasks, it is necessary to address the obstacles inherent to flow-through heaters. First, it has been determined that fluid flow rates of at least 2.25 gpm at 120.degree. F. are required to satisfy the lifestyle requirements of a typical residential application in most industrialized nations. Since water temperatures will range from 38.degree. F. to over 90.degree. F. depending upon the locale, source of water and season, the single source flow-through water heater requires a typical minimum fluid heating input capacity of at least 28,000 watts or approximately 95,000 BTU. In electric resistance type heating elements one should be concerned about watt density. High watt density elements have a much shorter service life than do low watt density elements. To obtain 28 KW and to maintain the benefits of a small heating systems, a water heater with multiple heating elements is preferred.
Much of the art before 1975 taught the use of fixed input heaters. Since domestic water heating applications require heating fluids at low (less than 1 gpm) to moderate (up to 3 gpm) flows, fixed input units had inherent limitations. In moderate flow conditions, there was sufficient water flow to absorb the heat and maintain a desirable temperature. At low flows, however, a high fixed input of heat would dangerously overheat the water, thereby creating potentially scalding conditions. The method of heating fluids in a fixed input, flow-through water heater thus limited the wattage and required flow activation devices that would prevent the activation of the heating elements until a safe minimum flow rate was achieved. Assuming the inlet water was 60.degree. F. or more, fixed input heaters having a total 28 KW heat input would be very dangerous at flows less than approximately 21/4 gallons per minute. For the foregoing reasons, electric fixed input flow-through water heaters were generally limited to 9 KW (approx. 30,000 BTU), and best served as single point-of-use heaters.
In addition to the above problems, additional considerations that should be addressed in the design and development of a commercially acceptable flow-through heater for domestic water heating are described below.
Flow-through water heaters with multiple heating elements are designed to have small heat exchangers with a relatively high wattage to fluid volume ratio. As shown in the prior art, the flowing water absorbs and carries off the heat satisfactorily during the heating mode, but at shutdown the latent heat in the heating elements will raise the water temperature to very high levels. This condition is aggravated in the sequential or staged activation controls schemes shown in prior art. In these designs, the heating elements, which are generally located in individual heating chambers, are sequentially activated with a first heating element energized to full power followed, depending upon demand, with a downstream element energized to full power and so on until all the elements necessary to produce the resulting desired temperature were continuously activated. At no flow, shutdown would occur and the elements would be deactivated. In some control schemes, shutdown would be affected by the sequential deactivation of the heating means, by first shutting down the last heating element to be activated and then sequentially shutting down the remaining elements until all were deactivated. In either scheme, at least one or more elements would have been on at full power and contain a significant amount of latent heat which was then transferred to the very small amount of static water in the heat exchanger or individual chamber. The results would be a very high overheating of the elements most energized at shutdown, and thus the overheating of the adjacent fluid. This over-temperature rise at shutdown can result in potential scalding conditions, particularly in point-of-use heaters where the distance from the heater to the fluid dispenser is very short.
The hotter the element, the hotter the water is heated, and the more minerals are precipitated out. This is particularly a problem in hard water areas. Thus another disadvantage of the cyclic overheating of water at shutdown is the resulting accelerated precipitation and accumulation of mineral deposits on the overheated elements and in the heat exchanger. Because of the relatively small chambers used in flow-through heaters, mineral deposits will quickly accumulate under these conditions, thereby shortening the life of the heating element and/or heat exchanger. In addition to the accumulation of mineral deposits in the water heater, these deposits may be carried out with fluid flow through the fluid distribution line and into the filter screens of appliances and fixtures, such as dishwashers, clothes washers, and faucets, thus undesirably increasing the maintenance for such appliances and fixtures.
One of the principal objectives of flow-through water heaters is to rapidly heat water on demand. The most common (devices used for temperature sensing in the and the response time for these devices to sense a temperature change may be in the range of two seconds. When one couples this response time to the time it takes to heat a resistance heating element and then to heat the water before the thermistor is heated, the overall time lag has been determined to be approximately 7 seconds. Accordingly, the control systems in prior art flow-through water heaters are of a hunting-type controller, which are subject to relatively high degrees of hysteresis in operating temperature. The volume of 120.degree. F. water required from a water heater to provide a normal shower, where the inlet water is 70.degree. F., is approximately 1.5 gpm. A 28 KW heater has the capacity to heat water flowing at 1.5 gpm through the water heater at approximately 2 degrees per second. The problems associated with a 7 second lag and the potential problems associated with overheating as described above are compounded in flow-through heaters having multiple high wattage heating means.
Almost everyone is familiar with the effects of the water temperature of a shower when a competing fixture is opened. When one is using a heated water supply source having relatively constant temperature, such as a storage tank heater, it is relatively easy to simply adjust the ratio of hot to cold water to resume desired temperature. This is not so with a flow-through water heater utilizing control systems having a high degree of hysteresis. The response time lags the temperature effects associated with rapid changes in flow rate, and annoying temperature swings result. Since pressure changes affect flow rate, the same annoying temperature swings apply to pressure changes. One commonly experiences pressure changes when using a private water well as the supply source. Even in community or city water supply systems, one can experience significant fluctuating pressures in high water demand periods.
When a flow-through water heater is serving multiple fixtures through normal runs of distribution piping, the piping serves to buffer temperature changes. When the heater overshoots temperature, part of the heat is absorbed by the piping and mixed with the flowing water to reduce the affects of small temperature swings. The same is true when the heater control system, as a result of hysteresis, undershoots the desired temperature. If the flow-through heater had been delivering water above set point, the pipes would have absorbed part of the heat and could be at a temperature above set point. When the heater then undershoots, the cooler water in part mixes with the hotter water and the excess heat in the piping is also transferred, in part, back to the water. This buffering effect can be beneficial in domestic water heater applications as piping runs from the heater to the fixture can easily be 30 ft or more.
Because power distribution systems are subject to high voltage transients and power surges, it is important for system reliability to provide protection to the triacs or other electronic solid state power control switching devices. Most commonly, this protection has been provided in the prior art by surge protectors having metal oxide varistors. These devices are wattage limited and are subject to destruction by large transients.
U.S. Pat. Nos. 5,216,743; 5,020,127; 4,604,515; 4,567,350; 4,511,790; 4,333,002; 3,952,182; and 3,787,729 recognize the benefits of multiple heating elements in a water heater, as well as the benefits of sequential or staged activation of the heating elements to provide better temperature control during operations. U.S. Pat. Nos. 5,216,743 and 5,020,127 included control systems which used sequential or staged modulation of the heating elements energized at zero crossing to reduce hysteresis in the temperature control as well as interference in the lighting circuits. U.S. Pat. No. 4,337,388, and European Patent EP 0 209 867 A2 recognize the need for better temperature control to avoid overheating and employed anticipation circuitry and modulating control. U.S. Pat. No. 3,952,782 and the cited European patent disclose initial venting of the heating chamber at or before start up of operation. U.S. Pat. No. 5,216,743 disclosed that gasses are produced during the heating process, and utilized continuous venting of the gasses to prevent damage to the heating elements or heating chamber, and to reduce the possibility of dangerous overheating during operations if the temperature sensor is deprived of fluid communication.
Most commercial therimistors reqiuire approximately 2 seconds to respond to full temperature change. This time may be referred to as the thermistors time constant. In a flow-through water heater wherein the thermistor is heated secondarily as a result of thermal lag, a system response delay occurs. The thermal lag is inherent since the heating element must be heated before heating the fluid to which the thermistor is responsive. In a flow-through water heater of the type disclosed herein, the system time constant required for the thermistor to respond to the full temperature change is approximately 7 seconds. Heating elements in prior art heaters are activated one at a time and sequentially, with a first heating element first being fully and continuously activated. The temperature change resulting from the activation of the first heating element is compared to reference set point voltages from which a demand signal is generated. The extent of reported demand is a function of the system temperature time constant. As the demand increases in relationship to the temperature changes and system time constant, indicating the need for additional heating, successive heating elements are sequentially activated from zero power to full continuous activation at 100%. Heating elements may be added until a sufficient number of elements are activated to achieve set point, as disclosed in U.S. Pat. No. 5,020,127. In most applications, fewer than the total number of elements are activated to achieve set point so that the duty cycle of the initially activated heating element is significantly greater than the final activated heating element. The activated heating elements are thus disproportionately hotter than in the areas in which the heating elements are not activated. As heating elements are de-energized in the reverse order to shutdown, hot spots occur as one or more elements are energized at full power in no flow conditions. Over time, the effects of this imbalance of heat distribution can damage the heat exchanger and significantly reduce the service life of the overworked heating elements. The resulting localized over-temperature resulting at shutdown will cause excessive precipitation of mineral deposits in hard water environments.
The disadvantages of the prior art are overcome by the present invention. An improved water heater is hereafter disclosed which utilizes a controller for more desirably regulating power to each of a plurality of heating elements. The techniques of the present invention are particularly well suited for use in a flow-through or tank-less water heater having multiple chambers therein each having one of the plurality of heating elements.